In my opinion, it is rather simple: if it is an acceptable work-around (for the asker, that is you), it can be an answer. If it is a work-around but you are asking it to get a better solution, it belongs to the question (not as a partial answer, but as a "this is what I've got, and this is why I don't like it").
The solution you propose indeed sounds more like a progress than an actual answer, but as said in the comments, if it works, there is no reason why it shouldn't be considered an answer. Maybe you should just make it clearer that it is a definitive answer rather than ongoing investigation, for example by adding some introduction like "Short version: run (...), or, if you don't have administrative privileges, (...)".
Also note that unanswered questions are a problem that all Q/A sites face and try to mitigate. As you can see, on tex.sx we aren't doing too bad, with only 6.3% of all questions left unanswered (4611 out of 72982). Part of the effort to reduce that number on the stackexchange network goes through some regular "Solve the unanswered" sessions, during which people methodically go through the list of unanswered questionsthe list of unanswered questions and try to provide an answer. Many of these unanswered questions will already contain a solution in the comments or in the question itself, in which case either the author of the solution will be pinged to make it an answer, or someone will move the solution into a community wiki answer. The point is your question would not have been left without an answer for more than a few months, and in particular not 10 years!